Xenotransplant Rejection: There is an ongoing shortage of human organs for transplant. This shortage has resulted in a long felt need for organs, and has resulted in attempts to develop xenotransplantation technology.
The primary non-primate candidate donor species for clinical xenotransplantation (e.g., the transplantation of non-human organs into human recipients) has been the pig. Swine provide an abundant supply of organs that are similar in size, anatomy, and physiology to their human counterparts (Auchincloss, 1988; Najarian, 1992; and Somervile and d'Apice, 1993). Transplantation of porcine pancreatic islets and of a pig liver into human patients has been reported, (Makowka, et al., 1993; Satake, et al., 1993; Tibell, et al., 1993), but the outcomes of these transplants need to be improved. One improvement that is needed is better control (e.g., inhibition) of transplant rejection.
The rejection of transplanted organs may involve both an extremely rapid hyperacute rejection (HAR) phase and a slower cellular rejection phase. HAR of discordant (i.e., non-primate) xenotransplants is initiated by preformed "natural" antibodies that bind to donor organ endothelium and activate complement attack by the recipient immune system (Dalmasso, et al., 1992; and Tuso, et al., 1993).
Activation of complement leads to the generation of fluid phase (C3a, C5a) and membrane bound (C3b and C5b-9, i.e., C5b, C6, C7, C8, and C9) proteins with chemotactic, procoagulant, proinflammatory, adhesive, and cytolytic properties (Muler-Eberhard, 1988). Immunohistological analysis of hyperacutely rejected xenotransplants reveals antibody deposition, complement fixation, and vascular thrombosis as well as neutrophil infiltration (Zehr, et al., 1994; Auchincloss, 1988; Najarian, 1992; Somervile and d'Apice, 1993; and Mejia-Laguna, et al., 1972).
While HAR is a major impediment to the xenotransplantation of vascularized organs, some discordantly xenotransplanted tissues (e.g., porcine pancreatic islets) do not appear to be rejected by this mechanism. Methods for the control of the HAR are also available. These include interference with the antibody antigen reactions responsible for initiating the HAR response, either by removing the antibodies from the circulation or by interfering with the expression of the antigens (see copending U.S. patent application Ser. No. 08/214,580, entitled "Xenotransplantation Therapies" and filed by Mauro S. Sandrin and Ian F. C. McKenzie on Mar. 15, 1994). Inhibition of complement attack on the xenotransplant may be accomplished by several means, including the use of complement inhibitors such as the 18 kDa C5b-9 inhibitory protein and monoclonal antibodies against human C5b-9 proteins as taught in U.S. Pat. No. 5,135,916, issued Aug. 4, 1992.
In order to better understand the porcine xenograft rejection phenomenon, studies have been undertaken to investigate interactions between human white blood cells and porcine cells, particularly porcine aortic endothelial cells (PAEC). The role of neutrophils in the actual destruction of xenografts has not been well characterized, and the precise mechanism of complement independent neutrophil activation and adherence to xenograft endothelium has heretofore been unknown.
Previous studies have shown that human complement component C3b (C3bi) deposited on PAEC mediates the binding of human neutrophils to the PAEC through interactions with the heterodimeric neutrophil cell surface receptor CD11b/CD18 (Vercellotti, et al., 1991). Furthermore, blocking HAR by inhibition or depletion of complement results in decreased neutrophil infiltration and increased xenograft survival, providing additional evidence for the role of complement in mediating human neutrophil binding to porcine endothelium.
However, a significant neutrophil infiltrate into PAEC monolayers has been observed even in the absence of complement activation (Leventhal, et al., 1993; and Pruitt, et al., 1991). The development of such infiltrates is believed to play an important role in xenograft rejection. Means and methods allowing the control or elimination of such interactions are thus needed in order to make the transplantation of porcine cells, tissues, or organs into human recipients more practicable.
E-selectin: E-selectin (also known as ELAM-1, CD62, and CD62E) is a cytokine inducible cell surface glycoprotein cell adhesion molecule that is found exclusively on endothelial cells. E-selectin mediates the adhesion of various leukocytes, including neutrophils, monocytes, eosinophils, natural killer (NK) cells and a subset of T cells, to activated endothelium (Bevilacqua, et al., 1989; Shimuzu, et al., 1991; Graber, et al., 1990; Carlos, et al., 1991; Hakkert, et al., 1991; and Picker, et al., 1991). The expression of E-selectin is induced on human endothelium in response to the cytokines IL-1 and TNF.alpha., as well as bacterial lipopolysaccharide (LPS), through transcriptional upregulation (Montgomery, et al., 1991).
Recently, the human leukocyte receptor for human E-selectin has been identified (Berg, et al., 1991 and Tyrrell, et al., 1991). This receptor contains sialic acid (sialyl Lewis x, and sialyl Lewis a) as a necessary component for interaction with the E-selectin protein.
Structurally, E-selectin belongs to a family of adhesion molecules termed "selectins" that also includes P-selectin and L-selectin (see reviews in Lasky, 1992 and Bevilacqua and Nelson, 1993). These molecules are characterized by common structural features such as an amino-terminal lectin-like domain, an epidermal growth factor (EGF) domain, and a discrete number of complement repeat modules (approximately 60 amino acids each) similar to those found in certain complement binding proteins.
Clinically, increased E-selectin expression on endothelium is associated with a variety of acute and chronic leukocyte-mediated inflammatory reactions including allograft rejection (Allen, et al., 1993; Brockmeyer, et al., 1993; Ferran, et al., 1993; and Taylor, et al., 1992). Other leukocyte-mediated inflammatory reactions associated with increased E-selectin expression on endothelium include delayed type hypersensitivity, immune complex-mediated lung injury, psoriasis, contact dermatitis, inflammatory bowel disease, and arthritis (Bevilacqua, et al., 1989; Bevilacqua and Nelson, 1993; Cotran, et al., 1L986; Koch, et al., 1991; Mulligan, et al., 1991; and Mulligan, et al., 1993).
Studies in which the expression of human E-selectin in cardiac and renal allografts undergoing acute cellular rejection was investigated have demonstrated that E-selectin expression is selectively upregulated in vascular endothelium of renal and cardiac tissue during acute rejection (Allen, et al., 1993; Brockmeyer, et al., 1993; Ferran, et al., 1993; and Taylor, et al., 1992). Additionally, increased E-selectin expression correlates with the early course of cellular rejection and corresponds to the migration of inflammatory cells into the graft tissue (Allen, et al., 1993; Brockmeyer, et al., 1993; Ferran, et al., 1993; and Taylor, et al., 1992). Taken together, these studies provide evidence that cytokine-induced expression of E-selectin by donor organ endothelium contributes to the binding and subsequent transmigration of inflammatory cells into the graft tissue and thereby plays an important role in acute cellular allograft rejection.
Endothelial cells have been shown to release a soluble form of E-selectin following in vitro activation (Pigott, et al., 1992; Newman, et al., 1993; and Leeuwenberg, et al., 1992). The demonstration of soluble E-selectin (sE-selectin) in the blood would therefore be taken as conclusive evidence of endothelial activation (Gearing and Newman, 1993). Table I shows the levels of sE-selectin found in healthy and sick patients in various studies, as reviewed in Gearing and Newman, 1993.
Elevated levels of sE-selectin have been found in diabetic patients independent of hypertension, nephropathy or renal failure, and whether or not they were insulin dependent (Gearing, et al., 1992 and Gearing and Newman, 1993). Similarly, sE-selectin was detected in patients with vasculidities including polyarteritis nodosum, giant cell arteritis, and scleroderma, with higher levels noted in lupus patients. Overall, however, there was no correlation with disease activity and only a weak correlation with the degree of organ involvement (Carson, et al., 1993). These studies suggest that the endothelium in these patients is activated, and that overt disease activity involves additional factors.
Soluble E-selectin levels have also shown marked elevations in sepsis, in one study a twenty-fold increase over the normal range (Newman, et al., 1993). In this and another study, the levels of sE-selectin appear to correlate with disease severity and/or outcome (Gearing and Newman, 1993). Higher levels or persistent elevation were associated with greater mortality and this is not unexpected in view of the widespread expression of E-selectin in most vessels in primates challenged with a lethal dose of live E. Coli (Redi, et al., 1991). Gram-negative as well as gram-positive infections seem to be associated with elevated levels. Patients with acute Plasmodium falciparum malaria have also shown elevated levels of sE-selectin (Hviid, et al., 1994, reviewed in Gearing and Newman, 1993).
There are two reports of sE-selectin in body fluids other than blood. One is from Dubois, et al., who detected sE-selectin in bronchioalveolar lavage (BAL) fluids from 16/50 patients with interstitial lung disease, but in only 2/16 control samples (Dubois, et al., unpublished, cited in Gearing and Newman, 1993). In the second, sE-selectin was recovered from BAL fluids of allergic subjects after segmental antigen challenge, but not saline challenge (Georas, et al., 1992).
Soluble E-selectin found in the blood is biologically active when measured by its ability to mediate adhesion of neutrophils to a surface. In the fluid-phase, recombinant soluble E-selectin can inhibit leukocyte adhesion (Lobb, et al., 1991). Furthermore, recombinant E-selectin, in either soluble or cell-surface bound forms, can activate the polymorphonuclear cell CD11b integrin receptor (Lo, et al., 1991 and Kuijpers, et al., 1991). The levels required for these effects are not found in the bloodstream, however such levels may be found at local sites of inflammation.
The mechanism of release of soluble E-selectin in vivo has not been established, but immunochemical evidence suggests a lost or defective cytoplasmic domain in the soluble E-selectin found in plasma. Unlike certain other cellular adhesion molecules, E-selectin does not appear to have an alternately spliced form lacking a transmembrane domain.